[github-actions]

Overview

Improves docs-vnext/models/catalog/model-choice-guide.mdx by consolidating repeated model-selection guidance into concise comparison and workload tables. Technical capabilities, reasoning trade-offs, latency guidance, use cases, pricing, Model Router, and Provisioned Throughput links remain intact.

Reduction

• Words: 1,247 to 630 (49.5% reduction)
• Lines: 95 to 54 (43.2% reduction)
• Diff: 35 insertions, 76 deletions

Changes

• Removed repetitive introductory descriptions of GPT-5 and GPT-4.1.
• Combined duplicated use-case bullet lists into one workload decision table.
• Condensed repeated latency explanations while preserving TTFT and TBT comparisons.
• Simplified table wording and retained all external resource links.

🗜️ Compressed by Documentation Unbloat
_{Comment /unbloat to run again}

• expires on Nov 16, 2026, 1:27 PM UTC